RSS Daily Digest

Fetch RSS feeds, parse articles, generate AI summaries, and compile a daily digest report in Markdown. Use when user asks for RSS summary, news digest, daily...

What to check before installing: - Clarify RSS_FEEDS_CONFIG: the skill metadata marks RSS_FEEDS_CONFIG as required/primary, but neither SKILL.md nor the scripts use it. Do not supply any secret or credential to that variable until the maintainer explains its purpose. - Curl is listed as required but unused: confirm whether the implementation expects to call curl or whether the requirement should be removed. - AI summarization is promised but not implemented: the provided scripts fetch and format articles but do not generate AI summaries or relevance scores. Ask the author whether summarization is performed by the agent (outside the scripts) or whether a missing module/step should be added; if you expect automatic LLM summaries, require an explicit implementation and clear disclosure of which model/service will be used and what credentials it needs. - Test in a safe environment: run the included scripts locally (they only fetch public RSS feeds listed in references/feed-sources.md) to confirm behavior and review the temporary file /tmp/openclaw-rss-articles.json and final digest outputs before giving this skill any elevated access. - If you need confidentiality for feed lists, avoid supplying secrets to RSS_FEEDS_CONFIG until its purpose is documented; store feed lists in the provided references/feed-sources.md or a secure location you control. Overall: the skill does not appear malicious, but the mismatches (unused env var, unused curl binary requirement, and missing summarization logic) are material and should be fixed or explained before trusting it in production.

Type: OpenClaw Skill Name: rss-daily-digest Version: 1.0.0 The OpenClaw skill 'rss-daily-digest' is designed to fetch RSS feeds, summarize articles, and generate a daily digest. All files (SKILL.md, Python scripts, feed-sources.md) align with this stated purpose. The Python scripts perform expected file I/O (reading local feed sources, writing to /tmp, outputting to stdout) and network requests to legitimate RSS feed URLs. There is no evidence of data exfiltration, malicious execution (e.g., `curl|bash`, `eval`), persistence mechanisms, or prompt injection attempts against the agent to deviate from its task. The instruction in SKILL.md to `pip3 install feedparser` is for a legitimate dependency and not indicative of malicious intent.